Instructions to use ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit")
model = AutoModelForCausalLM.from_pretrained("ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit

SGLang

How to use ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit with Docker Model Runner:
```
docker model run hf.co/ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit
```

open-llama-3b-opus-reasoning-sft-2k-4bit

Merged and 4-bit quantized version of open_llama_3b_v2 fine-tuned on reasoning data with <think> tags.

How it was made

Base: openlm-research/open_llama_3b_v2
QLoRA fine-tuned with LoRA adapter
Merged LoRA into base (16bit)
Quantized back to 4-bit NF4

Training Data

Crownelius/Opus-4.6-Reasoning-3300x (2,160 samples)
Roman1111111/claude-opus-4.6-10000x (9,633 samples)
11,669 samples after filtering (<= 2024 tokens)

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit")

messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "What is 10 * 5?"},
]
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True, enable_thinking=True)
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
output = model.generate(**inputs, max_new_tokens=512, temperature=0.7, do_sample=True)
print(tokenizer.decode(output[0], skip_special_tokens=False))

LoRA adapter

See ping98k/open-llama-3b-opus-reasoning-sft-lora-2k

Downloads last month: 8

Safetensors

Model size

4B params

Tensor type

F32

BF16

Model tree for ping98k/open-llama-3b-opus-reasoning-sft-2k-4bit

Base model

openlm-research/open_llama_3b_v2

Quantized

(30)

this model